Advance Ultravision - Blog: Halcon Deep Learning

Showing posts with label Halcon Deep Learning. Show all posts

Deep Learning with Densely Segmented Supermarket Dataset - MVTec Halcon

The Densely Segmented Supermarket (D2S) dataset is a benchmark for instance-aware semantic segmentation in an industrial domain. It contains 21,000 high-resolution images with pixel-wise labels of all object instances.

Find more information about Halcon here: https://mvtec.com

Find more information about Machine Vision here: http://advanceultravision.com/

Deep-Learning-Based Anomaly Detection - MVTec Halcon

Find more information about Halcon here: https://mvtec.com

Find more information about Machine Vision here: http://advanceultravision.com/

Anomaly Detection - MVTec Halcon

HALCON's deep-learning-based anomaly detection allows you to easily detect whether images contain deviations from known training data. To train this network, only a relatively low number of "good" images is required. Consequently, there is also no need to label any data.

Combined with its simple and flexible workflow, anomaly detection significantly increases the possibilities to solve your applications.

Find more information about Halcon here: https://mvtec.com

Find more information about Machine Vision here: http://advanceultravision.com/

Draw oriented rectangles with the MVTec Deep Learning Tool - MVTec Halcon

Many computer programs often offer the option to draw axis-aligned rectangles but not to draw oriented rectangles. Thus, we developed a method that enables you to label training data for deep learning efficiently with oriented rectangles.

This video shows you how to do this with the MVTec Deep Learning Tool. In this video, version 0.2 of the Deep Learning Tool is used.